Think Dataset: Filtered Institutional Books Published Before 1930

A filtered subset of the institutional-books-1.0 dataset containing English-language texts. The dataset applies filters for language and publication date, selecting books where the parsed year is before 1930. It was created by jbduran and last updated on 2026-06-04.

Use Cases

• Train language models on historical English prose based on the language filter.
• Analyze linguistic patterns in pre-1930 institutional publications based on the date filter.
• Study the evolution of vocabulary and grammar in formal English texts based on the source material.

Strengths

• Applies a specific language filter requiring English proportion >= 0.9.
• Focuses on a defined historical period with a parsed year < 1930 filter.

Limitations

• Row count, file formats, and column definitions are unknown, limiting suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.
• The dataset's redistribution is restricted, intended for private sharing only with authorized users.

Provenance

Source

Built from institutional/institutional-books-1.0.

Collection Method

Filtered based on language and publication date criteria.

Time Range

Publication years before 1930.

Freshness

Last updated 2026-06-04 18:11:13; freshness should be verified.